Abstract
This study aimed to evaluate the intra- and inter-instrument variability and reliability of the Axivity AX6 accelerometer under controlled technical conditions and human motion scenarios. In the first experiment, 12 accelerometers were affixed to a vibration platform and tested at four frequencies (2.2, 3.2, 6.5, and 9.4 Hz) along three axes to assess frequency- and axis-dependent variability. In the second experiment, four AX6 accelerometers were simultaneously placed on a subject's wrist and tested under four human motion conditions (walking at 4 km·h(-1) and 6 km·h(-1) and running at 8 km·h(-1) and 10 km·h(-1)). Results demonstrated low intra- and inter-instrument variability (CVintra: 3.3-4.5%; CVinter: 6.3-7.7%) with high reliability (ICC = 0.98). Similar results were observed in human motion conditions (CVintra: 5.3-8.8%; CVinter: 7.1-10.4%), with ICC values of 0.98 for combined devices, and 0.99 for each device individually. Despite statistically significant differences (p < 0.05) between devices in human motion all conditions, the variations remained below the minimal clinically significant difference threshold. These findings indicate that under technical conditions on a vibrating platform, and within the range of typical human accelerations, the Axivity AX6 is a reliable tool for measuring accelerations representative of physical activity. However, further research is necessary to validate its performance under free-living conditions.